The SCUBA Bright Quasar Survey II: unveiling the quasar epoch at submillimetre wavelengths

HAL Id: hal-00005493

https://hal.archives-ouvertes.fr/hal-00005493

Submitted on 16 Dec 2020

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

The SCUBA Bright Quasar Survey II: unveiling the

quasar epoch at submillimetre wavelengths

Robert S. Priddey, Kate G. Isaak, Richard G. Mcmahon, Alain Omont

To cite this version:

Robert S. Priddey, Kate G. Isaak, Richard G. Mcmahon, Alain Omont. The SCUBA Bright Quasar

Survey II: unveiling the quasar epoch at submillimetre wavelengths. Monthly Notices of the Royal

Astronomical Society, Oxford University Press (OUP): Policy P - Oxford Open Option A, 2003, 339

(4), pp.1183. �10.1046/j.1365-8711.2003.06279.x�. �hal-00005493�

The SCUBA Bright Quasar Survey (SBQS) – II. Unveiling the quasar

epoch at submillimetre wavelengths

Robert S. Priddey,

 Kate G. Isaak,

 Richard G. McMahon



and Alain Omont

1_{Astrophysics Group, Imperial College, Blackett Laboratory, Prince Consort Road, London SW7 2BZ} 2_{Cavendish Astrophysics, University of Cambridge, Cambridge CB3 0HE}

3_{Institute of Astronomy, Madingley Road, Cambridge CB3 0HA} 4_{Institut d’Astrophysique de Paris, CNRS, 98bis Bd. Arago, Paris, France}

Accepted 2002 November 8. Received 2002 October 22; in original form 2002 September 23

A B S T R A C T

We present results of the first systematic search for submillimetre (submm) continuum emission from z∼ 2, radio-quiet, optically-luminous (MB< −27.5) quasars, using the SCUBA array camera on the James Clerk Maxwell Telescope (JCMT). We have observed a homogeneous sample of 57 quasars in the redshift range 1.5< z < 3.0 – the epoch during which the comoving density of luminous active galactic nuclei (AGN) peaks – to make a systematic comparison with an equivalent sample at high redshift (z> 4; Paper I). The target sensitivity of the survey, 3σ = 10 mJy at 850 µm, was chosen to enable efficient identification of bright submm sources, suitable for detailed follow-up. Nine targets are detected with 3σ significance or greater, with fluxes in the range 7–17 mJy. Although the detection rate above 10 mJy is lower than that of the z> 4 survey, the weighted mean flux of the undetected sources, 1.9 ± 0.4 mJy, is similar to that at z> 4 (2.0 ± 0.6 mJy). The statistical significance of trends is analysed, and it is found that: (i) within the limited optical luminosity range studied, there is no strong evidence for a correlation between submm and optical luminosity; (ii) there is a suggestion of a variation of submm detectability with redshift, but that this is consistent with the K-correction of a characteristic far-infrared spectrum.

Key words: dust, extinction – quasars: general – galaxies: starburst – cosmology: observations

– submillimetre.

1 I N T R O D U C T I O N

By virtue of their high, sustained luminosity across the spectrum, quasars are ideal targets for studies of the evolution of structure throughout the history of the cosmos. Whilst a rare phenomenon in our local volume of space, the comoving density of luminous active galactic nuclei (AGN) rises sharply as one goes back in time, to a peak (Schmidt, Schneider & Gunn 1995; Fan et al. 2001) or a plateau (Miyaji, Hasinger & Schmidt 2000) around z= 2, an epoch depen-dence echoing the star formation history of galaxies (e.g. Madau et al. 1996; Steidel et al. 1999). It is widely believed that most massive galaxies undergo a short-lived AGN phase, as they build up the supermassive black holes observed in local galactic cores (Magorrian et al. 1998). Thus in observing a quasar, one catches its host galaxy at a significant period in its evolution.

E-mail: r.priddey@ic.ac.uk (RSP); isaak@mrao.cam.ac.uk (KGI); rgm

@ast.cam.ac.uk (RGM)

At far-infrared (FIR) to millimetre (mm) wavelengths, AGN have long been recognized as amongst the most prominent high-redshift sources [e.g. the compilations in McMahon et al. 1999 (M99); Rowan-Robinson 2000]. While there is strong evidence that emis-sion from this region of the spectrum is thermal re-radiation from warm dust (e.g. Hughes et al. 1993), it is not always clear that the AGN itself is the sole contributor to dust heating, and it is plausible that a fraction of the submillimetre (submm) luminosity could derive from stars in the surrounding galaxy. There would thus be a direct overlap between dusty quasars and SCUBA survey sources, which, it is argued, are the star-forming progenitors of massive spheroids. Hence, targeted submm and mm surveys of high-redshift AGN are valuable in elucidating the nature of cosmological submm sources, and the relation between the evolution of the host galaxy and its central black hole.

M99 observed a small sample of z> 4, radio-quiet quasars with the James Clerk Maxwell Telescope (JCMT) SCUBA submm cam-era to high sensitivity (σ850µm≈ 1.5 mJy), to determine the submm

properties of the ‘typical’ high-redshift AGN. A complementary,



1184

R. S. Priddey et al.

broad-but-shallow (σ850µm≈ 3 mJy) strategy was adopted by Isaak

et al. (2002, I02) – the SCUBA Bright Quasar Survey (SBQS) (Paper I) – with the principal aim of defining a statistically-significant sample of submm sources bright enough to permit a range of follow-up study. Roughly a quarter of the targets were brighter than the 3σ ∼ 10 mJy limit, suggesting that a substantial fraction of high-redshift quasars have FIR luminosities comparable to their blue luminosities. A natural question is whether this ubiqui-tous submm activity discovered at z> 4 is typical of high-redshift AGN in general: is the optically-luminous quasar phase always ac-companied by a dust-rich submillimetre source, or does this only occur at the highest redshifts? In the current paper we present the results of a comparative survey designed to address this question, targeting the ‘AGN epoch’ at z∼ 2, the era at which the space den-sity of quasars reaches its maximum, and by which most (>80 per cent) of the matter that will ever be accreted on to supermassive black holes has already served as fuel for AGN. A presentation of these results, along with brief analysis, is the purpose of the current paper: a more detailed and wide-ranging study, in the context of all the recent mm and submm quasar surveys, will be given in a forthcoming work (Priddey et al., in preparation).

We assume the currently favoured -dominated cosmology

M = 0.3, = 0.7, H0 = 65 km s−1Mpc−1 (). For

continu-ity with previous work, we will give alternatives in an Einstein– de Sitter (EdS) cosmology, M = 1,  = 0, with H0 =

50 km s−1Mpc−1.

2 S A M P L E S E L E C T I O N A N D O B S E RVAT I O N S Our aim was to find a sample of bright, medium-redshift quasars well-matched to the z> 4 sample observed by Isaak et al. (2002) (Fig. 1). To avoid too heterogeneous a sample, we restricted the in-put catalogues to a few large, homogeneous surveys. The final tar-get list comprises quasars preselected from the Large Bright Quasar Survey (LBQS; Hewett, Foltz & Chaffee 1995) and the Hamburg Quasar Survey (HS; Engels et al. 1998; Hagen, Engels & Reimers 1999). The first selection criterion is based on optical luminosity represented by absolute B-band magnitude (MB). At z = 2, the

Figure 1. Hubble diagram for quasar surveys used in the Paper I (Isaak et al.

2002, z> 4 selection) and the present work (z ≈ 2 selection). Note that in each case, the optical luminosity was designed to be MB< −27.5 in the EdS

cosmology for consistency with previous work. This is converted to the

cosmology employed in this paper, but it is clear that the difference between the samples is small. The black hole mass (right-hand axis) is calculated from MBassuming accretion at the Eddington rate.

J photometric band samples very close to rest-frame B, and H starts

to do so towards higher redshifts. Using a combination of these bands therefore gives an absorption-free assessment of the rest-frame optical magnitude, minimizing the error due extrapolation of the continuum. This overcomes many of the problems encountered when deriving MB for z> 4 quasars – e.g. contamination of the R band by strong emission and absorption features (as detailed in

Paper I). Similarly, at z≈ 2, observed-frame B becomes compro-mised by the strong CIVand Lyα emission lines.

Hence, to obtain a sample of luminous z≈ 2 quasars all with J and

H magnitudes, we cross-correlated catalogues of bright,

medium-redshift quasars with the 2MASS near-infrared catalogue through the web-based interface at IPAC (http://www.ipac.caltech.edu). Counting sources out to a radius of 60 arcsec enabled us to de-termine a maximum association radius of 5 arcsec, within which the probability of a chance association is<0.5 per cent. Absolute

B-band magnitudes were calculated from apparent J and H

magni-tudes thereby obtained (all our targets were detected in both bands) weighted according to the proximity of the band to B in the redshifted spectrum. They have been corrected for Galactic extinction, though in most cases this is negligible (<0.05 mag). We have assumed an optical spectral index αopt = −0.5 (where fν ∝ να), though, as

noted, the extrapolation error is small. Photometric errors in J and

H are typically 0.1–0.2 mag.

In order to match optical luminosity with the z> 4 sample, ob-jects with MEdS

B < −27.5 were initially selected – i.e. adopting the

same cosmology as per the selection of the z> 4 sample.1_{A wide}

redshift range 1.50< z < 3.00 was chosen to maximize the poten-tial number of targets (Fig. 1). To derive an observing sequence, the final selection was prioritized according to optical luminosity, the most luminous given preference.

The nominal completeness limit for 2MASS is 15.8 and 15.1 in

J and H, respectively. However, at the high Galactic latitudes

inhab-ited by these quasars, accurate detections up to 1 mag better than this can be achieved (Cutri et al. 2000). From this and our exten-sive quasar catalogue cross-correlations with 2MASS, we estimate that at z= 3, the 2MASS limit corresponds to MEdS

B ≈ −27.6, thus

for the greater extent of our redshift range we are not in danger of overlooking bright potential targets that failed to be detected by 2MASS.

The target list was correlated with the NRAO VLA Sky Survey (NVSS) to identify radio-loud sources, which were excised from the list. As discussed in M99 and I02, ‘radio loud’ in this context pri-marily refers to the extent of contamination of the thermal submm continuum by the extrapolated radio synchrotron. Assuming a spec-tral indexα = −0.5, a source with S1.4 GHz< 1.5 mJy (below the

NVSS limit) will have S850µm< 0.1 mJy.

The JCMT observing strategy for this project was the same as for the comparable z> 4 program described in detail in I02. To reiterate: the required RMS (1σ = 3 mJy) can be reached in a short amount of observing in relatively poor (zenith trans-mission 65 per cent) weather, so the JCMT fallback queue was utilized. SCUBA was employed in photometry mode, that is, the source is placed on the central bolometer, while the rest of the array samples the sky. Flux calibration was determined from the planets Mars and Uranus, or from secondary continuum standards

1_{Note that this selection was made in the EdS cosmology; as shown in Fig. 1} and Table 1, converting to the cosmology introduces an average shift of −0.2 mag. Most importantly, the overall difference between the z = 2 and

z> 4 samples is small.

C

2003 RAS, MNRAS 339, 1183–1188

Submillimetre properties of z

∼ 2 quasars

1185

inter-face http://www.ast.cam.ac.uk/∼apmcat) was used to verify coor-dinates, which are given in Table 1.

3 R E S U LT S A N D C O M M E N TA R Y

A total of 57 z∼ 2 quasars have been observed with SCUBA in the current survey, nine of which are detected with>3σ significance at 850µm. The median RMS flux isσ850= 2.8 mJy, with a narrow

(0.4 mJy) interquartile range. Four of the sources have 3σ > 10 mJy: in the following analysis, we exclude these from our complete statistical sample, which thus consists of 53 sources.

One detection, the z= 2.6 quasar LBQS B0018−0220, is signif-icantly brighter than the nominal 10-mJy limit: with S850 = 17 ±

3 mJy, this is an exceptionally bright source, comparable to some of the brightest high-redshift submm sources known. To recover most of the other detections, one must go down to≈8 mJy. In contrast, seven of the eight detections from the z> 4 SCUBA Bright Survey (I02) lie above 10 mJy. In Fig. 2, 850-µm fluxes of the z∼ 2 sources reported in this paper are plotted along with the z> 4 sample de-scribed in I02. The curves represent the flux one would observe from an object of fixed luminosity as a function of redshift, assuming it has the mean FIR SED determined from z> 4 quasars by Priddey & McMahon (2001, PM01), characterized by an isothermal temper-ature and emissivity index T = 40 K and β = 2, respectively. The large K-correction towards high redshift is one possible reason why the z> 4 sources appear systematically brighter, however it is not obvious that the same SED is valid for all objects at all redshifts. (Indeed, one of the motivations for this project was to yield a sample bright enough to be studied at a number of submm wavelengths, so that the typical FIR SED of a z= 2 quasar could be determined.)

4 S TAT I S T I C A L A N A LY S I S 4.1 Submillimetre versus optical

Fig. 3 shows 850-µm fluxes plotted against absolute magnitudes for the z≈ 2 sample. (NB: we do not correct the optical for intrinsic absorption.) Any correlation between submm and optical is even less evident than in the case of the z > 4 sources of I02. This can be demonstrated formally using statistical tests. If there were a correlation between detectability and absolute magnitude, we might expect the detected sources to be distributed with respect to MBin

a different way than the whole sample. Applying the Kolmogorov– Smirnov (K–S) test informs us that the null hypothesis that the distributions are the same. Note that, unlike the z> 4 objects, this material is less prone to systematic error in the determination of MB:

the magnitudes were calculated homogeneously, and the need for continuum extrapolation was minimized. Yet the scatter remains, which suggests that it derives from an underlying variance in the submm–optical relation.

As discussed in I02, one might naively expect a submm– optical correlation, whether the dust-heating source is stars (because spheroid mass scales with black hole mass) or the AGN (because both LFIRand LBscale with bolometric luminosity). There is much

scope for complexity which would smear out any correlation, for example the relative timing between AGN fuelling and starburst, or, in an AGN-powered scenario, the effects of varying the dust-torus geometry. As a caveat, note that the nominal detection threshold, 10 mJy, corresponds (at z= 2) to LFIR≈ 2 × 1013L ≈ νLB for

the value of the magnitude cut, MB= −27.5. We are therefore

prob-ably sampling only the bright tail of the FIR luminosity distribution,

and we are biased towards sources lying just above this LFIR≈ νLB

threshold. Thus, for the detected sources, any such relation should be derived with caution.

4.2 Optical colour

As described in Section 2, 2MASS J and H magnitudes were ob-tained primarily to provide a homogeneous optical luminosity scale, directly sampling rest-frame B. However, we can use them to as-sess, roughly, whether there is any correlation between submm de-tectability and optical colour – for example, whether submm excess is connected with intrinsic reddening. The mean B− H colour of the sample is 1.9± 0.5, whereas that of the detections is 2.0 ± 0.5; similarly, B− J of the sample is 1.3 ± 0.4, of the detections 1.4 ± 0.5. There is no evidence, then, that the detected quasars are redder than the underlying sample. However, given the large uncertainties, further work is required to address the question systematically.

Note that while we are requiring our sample to be bright in rest-frame B, the selection procedure for these surveys (e.g. LBQS

BJ< 18.75) also requires these objects to be bright in the rest-frame

ultraviolet (UV). This limit in observed B translates into a limit on redness for a given luminosity at a given redshift. In the very worst case (z= 3, MB = −27.5), objects redder than α = −0.7 are

ex-cluded. On the other hand, by selecting objects bright in 2MASS, we are biased against very blue quasars. There is, however, no sta-tistically significant correlation between colour and either MBor z;

and the distribution of spectral indices for the sample is similar to (if a little narrower than) that of the whole LBQS (medianα = −0.3, Francis et al. 1991).

4.3 Broad absorption line quasars

Omont et al. (1996) suggested that those quasars exhibiting broad absorption lines (BALs) in their optical spectra are preferentially detected in the submm. This is based on the marginal evidence that two out of their six z> 4 IRAM detections were BALs, rela-tive to the background BAL abundance of≈10 per cent (Weymann 1991). We are now in a better position to test this hypothesis. Con-sidering sources from the current paper, one of the detections (HS B1141+4201) is unambiguously a BAL, another (HS B1310+4308) is possibly a weak BAL. In comparison, four of the non-detections are BALs, with a number of others showing signs of weak absorp-tion, a fraction consistent with the expected 10 per cent. Thus, al-though far from providing proof, these data do not rule out the Omont et al. hypothesis. In a forthcoming paper (Priddey et al., in preparation), we shall discuss, in detail, the evidence from a statis-tically significant combination of all recent submm and mm quasar surveys.

4.4 Submillimetre versus redshift

Fig. 4 is a cumulative histogram of redshift for the z∼ 2 sample and for the detected subsample. The distributions appear different to the eye, with most of the detections lying at higher redshift. The median redshift of the sample is z= 2.3, and that of the detections is z = 2.6. The maximum difference between the distributions of the detected and the parent samples is 0.42, corresponding to a K–S level of 5 per cent significance, whilst for the the detected and undetected samples, the maximum is 0.51, giving a≈1 per cent level of significance. Within this limited redshift range, therefore, there is a suggestion that the submm detection rate increases with redshift over the range 1.5< z < 3.0. In Fig. 5, the FIR luminosity as a function of redshift



1186

R. S. Priddey et al.

Table 1. APM astrometry, and 2MASS NIR and SCUBA submm photometry of z∼ 2 quasars.

Target name RA Dec. z J H M_B(MEdS

B ) Observation S850± σ850 Notes (J2000) (J2000) dates (mJy) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) LBQS B0018−0220a _{00 21 27.37 −02 03 33.8 2.56 16.2 15.4 −28.6(−28.3) 05/08/00 17.2± 2.9} HS B0035+4405a _{00 37 52.31 +44 21 32.9 2.71 15.9 15.4 −28.7 (−28.4) 22/09/00 9.4± 2.8} HS B0211+1858 02 14 29.71 +19 12 37.6 2.47 16.2 15.5 −28.3 (−28.0) 22/09/00 7.1± 2.1 HS B0810+2554b _{08 13 31.30 +25 45 02.9 1.50 14.1 13.2 −29.2 (−29.1) 07, 09, 10/08/00 7.6± 1.8} HS B0943+3155 09 46 23.21 +31 41 30.5 2.79 16.5 16.0 −28.1 (−27.8) 08/10/00 9.6± 3.0 HS B1140+2711 11 42 54.27 +26 54 57.8 2.63 15.8 15.2 −28.8 (−28.5) 17/01/01 8.6± 2.6 HS B1141+4201 11 43 52.04 +41 45 19.8 2.12 15.5 15.1 −28.5 (−28.3) 18/01/01 8.6± 2.6 BAL HS B1310+4308 13 12 48.73 +42 52 36.8 2.60 16.0 15.8 −28.2 (−27.9) 13/04/01 10.0± 2.8 weak BAL HS B1337+2123 13 40 10.84 +21 08 44.5e 2.70 16.5 15.6 −28.4 (−28.1) 13/04/01 6.8± 2.1 LBQS B0009+0219 00 12 19.64 +02 36 35.4 2.64 16.3 16.0 −28.0 (−27.7) 06/08/00 1.4± 3.2 LBQS B0009−0138 00 12 10.91 −01 22 07.7 2.00 16.2 15.7 −27.6 (−27.4) 06/08/00 2.2± 2.8 LBQS B0013−0029 00 16 02.41 −00 12 25.2 2.08 16.3 15.2 −27.7 (−27.5) 06/08/00 3.2± 3.3 HS B0017+2116 00 20 10.85e _{+21 32 51.4}e _{2.02 16.0 15.5 −27.8 (−27.6) 22/09/00 2.3± 3.0} LBQS B0025−0151 00 27 33.82 −01 34 52.4 2.08 16.3 15.7 −27.7 (−27.5) 06/08/00 4.8± 2.47 BAL HS B0029+3725c 00 32 10.08 +37 42 32.5 1.85 15.8 15.5 −28.6 (−28.4) 14/09/00 2.2± 2.6 HS B0036+3842 00 39 07.49 +38 59 15.5e 2.36 16.4 15.6 −28.0 (−27.7) 22/09/00 3.0± 3.0 BAL HS B0037+1351 00 40 23.76 +14 08 07.5 1.87 15.6 15.1 −28.1 (−27.9) 06/08/00 3.4± 3.1 HS B0042+3704 00 44 48.3f +37 21 14f 2.41 16.4 15.9 −28.0 (−27.7) 12/10/00 −1.2 ± 2.6 HS B0105+1619 01 08 06.47 +16 35 50.4 2.64 15.7 15.1 −28.9 (−28.6) 04/08/00 3.8± 2.6 HS B0119+1432 01 21 56.06 +14 48 24.0 2.87 15.5 15.1 −29.0 (−28.7) 04/08/00 3.6± 2.6 HS B0150+3806 01 53 13.55 +38 21 24.2 1.96 16.0 15.4 −27.8 (−27.6) 06/08/00 2.2± 3.0 HS B0202+1848 02 05 27.52 +19 02 29.8 2.70 15.7 15.3 −28.8 (−28.5) 05/08/00 0.6± 3.3 HS B0218+3707 02 21 05.52 +37 20 46.2 2.41 15.8 14.9 −28.6 (−28.4) 04/08/00 1.9± 2.6 HS B0219+1452 02 22 31.71e _{+15 06 28.6}e _{1.71 14.8 14.2 −28.8 (−28.6) 05/08/00 −5.3 ± 2.8} HS B0248+3402 02 51 27.78 +34 14 42.1 2.23 15.4 15.0 −28.6 (−28.4) 04/08/00 −0.2 ± 2.7 HS B0752+3429 07 55 24.10 +34 21 34.2 2.11 16.1 15.4 −27.9 (−27.7) 22/09/00 2.4± 2.8 HS B0800+3031 08 03 42.05 +30 22 54.8 2.02 15.1 14.5 −28.8 (−28.6) 07/07/00 −3.0 ± 3.4 HS B0808+1218 08 10 56.9f _{+12 09 14}f _{2.26 16.2 15.5 −28.0 (−27.8) 22/09,08/10/00 1.1± 2.1} HS B0821+3613 08 25 07.66 +36 04 11.5 1.58 15.5 14.7 −27.9 (−27.7) 22/09,08/10/00 1.5± 2.0 HS B0830+1833 08 32 55.63 +18 23 00.7 2.27 15.9 15.4 −28.3 (−28.1) 08/10/00 −0.5 ± 3.1 HS B0834+1509 08 37 12.87e _{+14 59 17.5 2.51 16.2 15.4 −28.3 (−28.0) 08/10/00 −1.0 ± 2.8} HS B0926+3608 09 29 52.14 +35 54 49.8 2.14 16.3 15.5 −27.7 (−27.5) 08/10/00 1.2± 2.8 HS B0929+3156 09 32 08.77 +31 43 28.0e 2.08 16.1 15.6 −27.9 (−27.7) 08/10/00 5.7± 3.0 weak BAL HS B0931+2258 09 34 42.26 +22 44 39.5 1.74 15.4 14.8 −28.2 (−28.0) 08/10/00 −3.5 ± 2.8 HS B1002+4400 10 05 17.47 +43 46 09.3 2.08 15.5 15.0 −28.5 (−28.3) 13,19/04/01 8.2± 2.9 HS B1031+1831 10 34 28.89 +18 15 32.4 1.53 15.1 14.3 −28.3 (−28.1) 19/04/01 −0.1 ± 2.7 BAL HS B1049+4033 10 51 58.71e +40 17 37.0e 2.15 15.7 15.1 −28.4 (−28.2) 19/04/01 3.9± 3.2 HS B1111+4033 11 13 50.94e _{+40 17 21.5 2.18 16.0 15.7 −28.1 (−27.9) 19/04/01 2.2± 2.8} HS B1115+2015 11 18 00.52e _{+19 58 53.8}e _{1.93 15.5 15.0 −28.4 (−28.2) 19/04/01 5.3± 2.8 BAL} HS B1126+3639 11 28 57.84e _{+36 22 50.3}e _{2.89 16.3 15.8 −28.3 (−28.0) 19/04/01 −2.5 ± 2.8} HS B1155+2640 11 57 41.91 +26 23 56.2 2.80 16.6 15.8 −28.3 (−28.0) 19/04/01 3.5± 2.7 HS B1200+1539 12 03 31.28e _{+15 22 54.4}e _{2.97 15.8 15.2 −28.9 (−28.6) 08/01/01 −5.1 ± 3.1} LBQS B1210+1731d _{12 13 03.02 +17 14 23.4}e _{2.54 16.3 15.6 −28.3 (−28.0) 08/01/01 1.3± 2.8} HS B1215+2430 12 18 10.98 +24 14 10.8 2.36 15.8 15.1 −28.6 (−28.3) 19/04/01 6.0± 2.7 HS B1302+4226 13 04 25.56 +42 10 09.7 1.91 15.3 14.7 −28.5 (−28.3) 18/01/01 2.4± 2.4 weak BAL HS B1326+3923 13 28 23.73e _{+39 08 17.8 2.32 15.4 14.8 −28.8 (−28.6) 08/01/01 7.4± 3.0} LBQS B1334−0033 13 36 47.16 −00 48 57.4 2.80 16.3 15.6 −28.0 (−27.8) 08/01/01 2.5± 2.6 HS B1356+3113 13 59 08.39 +30 58 30.8 2.26 16.2 15.7 −28.6 (−28.3) 20/03/01 3.2± 3.0 HS B1417+4722 14 19 51.84 +47 09 01.1 2.27 15.7 15.1 −28.5 (−28.3) 18/01/01 8.8± 3.4 HS B1422+4224 14 24 35.96 +42 10 30.6e _{2.21 16.2 16.0 −27.7 (−27.5) 19/04/01 10.7± 4.7} HS B1703+5350 17 04 06.74 +53 46 53.6 2.37 15.9 15.3 −28.4 (−28.1) 20/03/01 −0.1 ± 2.6 HS B1754+3818 17 56 39.6f +38 17 52f 2.16 16.0 15.4 −28.1 (−27.9) 20/03/01 0.7± 2.5 HS B2134+1531 21 36 23.86e +15 45 08.4 2.13 15.6 14.8 −28.4 (−28.2) 04/08/00 −0.2 ± 2.9 LBQS B2244−0105 22 46 49.30 −00 49 53.9 2.03 15.9 15.7 −27.9 (−27.7) 14/09/00 0.2± 2.8 HS B2245+2531 22 47 27.40 +25 47 30.5 2.15 15.4 14.9 −28.5 (−28.3) 04/08/00 3.5± 2.6 HS B2251+2941 22 53 38.59 +29 57 12.5 1.57 15.5 14.7 −27.9 (−27.7) 14/09/00 −1.2 ± 2.6 HS B2337+1845 23 39 44.77e _{+19 01 51.1 2.62 15.6 14.9 −29.1 (−28.8) 04/08/00 2.4± 4.0} a_{Detected at 450µm (Isaak et al., in preparation);}b_{IRAS source, S}

60= 284 ± 40 mJy, S100= 538 ± 160 mJy. Recent HST STIS imaging shows this quasar to be quadrupally lensed (Reimers et al. 2002);c_S

1.4 GHz= 2.6 mJy;dS1.4 GHz= 2.0 mJy.eAPM position offset from published coordinates by 1 arcsec or more: 2MASS position favours APM.f_{Blending in APM, published (HS) coordinates used.}

C

2003 RAS, MNRAS 339, 1183–1188

Submillimetre properties of z

∼ 2 quasars

1187

Figure 2. Flux at 850µm against redshift, plotted for SBQS detections

from Paper I (I02, open symbols) and the current work (filled symbols). The dotted line represents the nominal 3σ = 10 mJy limit for this work, the approximate threshold for viable follow-up such as CO line detection. The two curves are, for two different cosmologies, the flux one would detect from a source of fixed luminosity having the mean isothermal SED of PM01 (T= 40 K, β = 2). They are arbitrarily normalized, and are plotted only to illustrate the necessary K-correction between the redshift ranges of the two surveys.

Figure 3. 850-µm flux against absolute B magnitude ( cosmology) for

the complete z≈ 2 SCUBA sample. Submm detections are plotted as large solid squares with error bars reflecting photometric errors in S850and near-infrared magnitudes. Non-detections are plotted by thin lines, as upper limits corresponding to signal+90 per cent confidence, with a vertical error bar terminating at the value of the signal. The weighted means of the non-detections in each of four magnitude bins are also plotted as open symbols.

has been derived by calculating theσ−2-weighted mean of all the SBQS data – I02 for z> 3, and the current work for z < 3 – and for detections and non-detections alike. The 850-µm flux has been converted to a luminosity assuming the isothermal SED derived in PM01. Changing this assumption would shift the z> 4 and z ∼ 2 points relative to one another; (lowering the temperature orβ

increases the relative luminosity of the z> 4 points). In I02, the

fluxes of the non-detections were stacked to give an estimate of the

Figure 4. Cumulative histogram for redshifts of the z≈ 2 SCUBA sample.

The difference between the distributions of the detected (solid) and unde-tected (dotted) and parent (dashed) samples is clear to the eye, and supported by formal statistics.

Figure 5. FIR luminosity for SBQS quasars (crosses) as a function of

redshift, determined from the weighted mean of all data within each bin. Also plotted, for comparison, are the radio galaxy points (squares) and the power-law fit from Archibald et al. (2001).

submm flux of an average z> 4 quasar – after checking that sky-subtraction had been performed effectively, by observing that the distribution of the signal in all off-source pixels is a Gaussian with zero mean. We repeat the experiment with the current z∼ 2 data, and obtain S= 1.9 ± 0.4 mJy, which agrees within the uncertainties with the z> 4 value, 2.0 ± 0.6 mJy.

We can compare these results with those drawn from the SCUBA survey of high-redshift radio galaxies (Archibald et al. 2001), from which a dramatic increase of FIR luminosity with redshift was inferred, LFIR ∼ (1 + z)3−4, over the whole redshift range 1<

z< 4 (see Fig. 5). While our tentative low-redshift decline echoes

this behaviour, the difference between z> 4 and z = 2 is far less marked. It is necessary to improve the statistics and to consider the selection effects before drawing firm conclusions. Nevertheless, the disparity between the two populations is a potentially revealing commentary on the nature of radio loudness in AGN. Conceivably, radio galaxies follow a more dramatic evolution than the majority of AGN – of which they are perhaps the most extreme members – their radio-loudness originating from a difference in formation



1188

R. S. Priddey et al.

mechanism. Note however that the masses of their central black holes may be no more extreme than those of the most optically-luminous radio-quiet quasars in the present sample, some of which would have M > 109_M

, even assuming accretion at the maxi-mum (Eddington) rate.

The present results show that, on average, the submm properties of luminous, radio-quiet quasars at z∼ 2 are comparable, within the uncertainties, to those at z> 4, despite the lack of brighter sources in the present sample. Although the uncertainties are too great to draw detailed conclusions, one can speculate on the physical pro-cesses underlying the redshift variation of submm detectability at fixed optical luminosity. In a starburst scenario, for example, the gas accretion efficiency and the star formation efficiency could each depend in a different way upon redshift (e.g. through the dynamical time: Kauffman & Haehnelt 2000). At z> 4, the host galaxy is presumably gas-rich (and dust-rich), whereas by z< 2 most of the gas has formed into stars. At yet higher redshifts (z> 5) the youth of the universe may preclude the production of enough obscuring dust for the quasar to be a luminous submm source (Priddey et al., in preparation). Expanding the redshift range of Fig. 5 in each direc-tion is a project currently in progress, but at present it is intriguing that the average submillimetre loudness seems to decline when the population itself declines.

4.5 Caveat: gravitational lensing

A factor that we have not addressed in deriving this result is the influence of gravitational lensing. In common with many of bright submm sources found in surveys, it is likely that fluxes are boosted by lensing. This is likely to lead to bias, as the lensing optical depth increases with redshift (e.g. Barkana & Loeb 2000). Thus the z> 4 sources are likely to be intrinsically fainter than they appear, relative to those at z= 2. Considering individual sources, HST STIS imaging (after the current data were obtained) showed HS0810+2554 to be lensed into a quadruple source with a tight (<1 arcsec) separation between the components (Reimers et al. 2002). This explains the extremely high infrared luminosity implied by its detection in the

IRAS Faint Source Catalogue.

We stress that no corrections for lensing have been applied in the current work, for this is complex and model-dependent. However, this important issue will be tackled in quantitative detail in a future paper (Priddey et al., in preparation), where we present a detailed statistical analysis of all recent JCMT/SCUBA and IRAM/MAMBO high-redshift quasar data.

5 S U M M A R Y A N D F U T U R E W O R K

In this paper, we have presented the first results from a targeted SCUBA survey of optically luminous (MB < −27.5), radio-quiet

quasars at z∼ 2. This is a continuation of the SCUBA Bright Quasar Survey whose preliminary results, at z> 4, were reported by Isaak et al. (2002). The present data confirm the presence of a large scatter in the correlation between submm and optical luminosity, despite having minimized the errors in measurement of the latter. Compar-ing the z> 4 and z < 3 data sets shows that there is no evidence for a variation of submm properties of luminous quasars across this

redshift range, once one has allowed for a K-correction appropriate for cool, isothermal dust. However, there is a suggestion that the characteristic submm luminosity increases with redshift between

z= 1.5 and z = 3.

In forthcoming papers, we shall present multiwavelength follow-up of bright sources from the present sample, results of comparative studies to improve the redshift coverage, detailed statistical analysis of>200 high-redshift quasars observed at (sub)mm wavelengths, and astrophysical interpretation of the findings.

AC K N OW L E D G M E N T S

For support through the period during which the bulk of this work was carried out, RSP and KGI thank PPARC and RGM thanks the Royal Society. We are grateful to the JACH staff and those JCMT observers, their projects displaced by poor weather, who gathered data for us in fallback mode. The JCMT is operated by JAC, Hilo, on behalf of the parent organisations of the Particle Physics and Astron-omy Research Council in the UK, the National Research Council in Canada and the Scientific Research Organisation of the Netherlands. We thank the anonymous referee for constructive comments. R E F E R E N C E S

Archibald E. N., Dunlop J. S., Hughes D. H., Rawlings S., Eales S. A., Ivison R. J., 2001, MNRAS, 323, 417

Barkana R., Loeb A., 2000, ApJ, 531, 613

Cutri R. M. et al., 2000, Explanatory Supplement to the 2MASS Second Incremental Data Release. http://www.ipac.caltech.edu/2mass/ releases/second/doc/explsup.html

Engels D., Hagen H.-J., Cordis L., Koehler S., Wisotzki L., Reimers D., 1998, AAPS, 128, 507

Fan X. et al., 2001, AJ, 121, 54

Francis P. J., Hewett P. C., Foltz C. B., Chaffee F. H., Weymann R. J., Morris S. L., 1991, ApJ, 373, 465

Hagen H.-J., Engels D., Reimers D., 1999, AAPS, 134, 483 Hewett P. C., Foltz C. B., Chaffee F. H., 1995, AJ, 109, 1498

Hughes D. H., Robson E. I., Dunlop J. S., Gear W. K., 1993, MNRAS, 263, 607

Isaak K. G., Priddey R. S., McMahon R. G., Omont A., Cox P., Peroux C., Sharp R., Withington S., 2002, MNRAS, 329, 149 (I02)

Kauffmann G., Haehnelt M., 2000, MNRAS, 311, 576

McMahon R. G., Priddey R. S., Omont A., Snellen I., Withington S., 1999, MNRAS, 309, L1 (M99)

Madau P., Ferguson H., Dickinson M., Giavalisco M., Steidel C., Fruchter A., 1996, MNRAS, 283, 1388

Magorrian J. et al., 1998, AJ, 115, 2285

Miyaji T., Hasinger G., Schmidt M., 2001, A&A, 369, 49

Omont A., McMahon R. G., Cox P., Kreysa E., Bergeron J., Pajot F., Storrie-Lombardi L. J., 1996, A&A, 315, 1

Priddey R. S., McMahon R. G., 2001, MNRAS, 324, L17 (PM01) Reimers D., Hagen H.-J., Baade R., Lopez S., Tytler D., 2002, A&A, 382,

L26

Rowan-Robinson M., 2000, MNRAS, 316, 885

Schmidt M., Schneider D. P., Gunn J. E., 1995, AJ, 110, 68

Steidel C., Adelberger K., Giavalisco M., Dickinson M., Pettini M., 1999, ApJ, 519, 1

Weymann R. J., Morris S. L., Foltz C. B., Hewett P. C., 1991, ApJ, 373, 23

This paper has been typeset from a TEX/LA_{TEX file prepared by the author.}

C

2003 RAS, MNRAS 339, 1183–1188